Due to the severity of the patient's condition we had to use allogenic dog adipose derived MSC for the generation of biologically active 3D matrix. MSCs were isolated using standard techniques as described previously (13). Briefly, the adipose tissue of donor dogs' greater omentum was isolated in a veterinary operating room during ovariohysterectomy under sterile conditions. The tissue was transported in a sterile flask with sterile 0.9% NaCl for further work in cell culture laboratory (within 2 h). The adipose tissue was cut to pieces of about 1 cm² in a laminar flow biosafety cabinet under sterile conditions. Blood cells were washed out using centrifugation at 500 × g for 5 min. Adipose tissue was then incubated with the sterile solution of crab hepatopancreas collagenase (Biolot, Russia) to a final concentration of 0.2% for an hour at 37°C with shaking at 200 rpm. Next, the homogenate was centrifuged for 5 min at 500 × g, the enzymatic solution was then decanted and the cell pellet containing the stromal-vascular fraction was suspended in Dulbecco's phosphate-buffered saline (DPBS) solution and centrifuged twice for 5 min each at 500 × g to remove any residual enzymes. The obtained cells were cultured in α-MEM medium with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine (all from PanEco, Russia). The culture medium was changed every 3 days. Immunophenotyping of MSCs was performed using antibodies CD10 FITC, CD71 FITC (Sorbent, Russia), CD34 AF488, CD45 AF488, CD105 AF488 (Biolegend), CD44, stro-1, and Thy-1 (Santa Cruz, USA) according to manufacturers' protocols. Results were evaluated using a confocal fluorescence scanning microscope LSM 780 (ZEISS, Germany).

Recombinant replication-defective adenoviruses Ad5-VEGF165 and Ad5-FGF2 were obtained using Gateway cloning technology as reported previously (14). Vegf165 gene cDNA fragments were amplified using thermocycler C1000 Thermo Cycler (BioRad), Phusion High fidelity DNA Polymerase (Finnzymes, Thermo Fisher Scientific, USA), and specific primers (Synthol, Russia). The purified PCR amplification products were cloned into a plasmid vector pENTR-D/TOPO (Invitrogen Thermo Fisher Scientific, USA) using topoisomerase followed by transformation into competent E. coli Top 10 cells. PCR screening of the colonies was carried out using vector-specific primers. Target recombinant plasmid uptake was confirmed by sequencing and restriction analysis.

PCR amplification of fgf2 gene fragments was carried out in two stages. Initially attB—sites were connected using gene-specific primers hFGF2—attB1 and hFGF2—attB2 (15). The second stage was to increase the length of the nucleotide att-sites sequence which was performed using the adapter primer GW-attB1 and GW-attB2 (Liteh, Russia). BP- recombination was performed according to a standard protocol (Invitrogen Thermo Fisher Scientific, USA) followed by transformation into competent cells E. coli Top 10. PCR screening of colonies and target recombinant plasmid uptake confirmations were carried out as described above.

LR-recombination of gene cDNA from donor plasmids pENTR- VEGF165 and pDONR-FGF2 into vectors plasmid pAd/CMV/V5-Dest (Invitrogen Thermo Fisher Scientific, USA) was performed to create adenoviral expression constructs. Transformation of competent E. coli cells, the PCR screening of colonies and target recombinant plasmid receiving confirmation were carried out as described above.

To produce recombinant adenovirus Ad5-VEGF165 and Ad5—FGF2, adenoviral plasmid vector was transferred into a linear form using a restriction enzyme PacI. HEK293A cells were transfected with purified linear plasmids using transfection reagent TurboFect. On the 10th day after transfection, fields of cytopathic effect became apparent and the cell suspension was collected. Cells were cryolysed, cell debris was eradicated using centrifugation to prepare a crude viral lysate. Amplification of recombinant adenoviruses Ad5-VEGF165 and Ad5—FGF2 was performed in HEK293A cell cultures. Adenoviruses were concentrated using ultracentrifugation in cesium chloride density gradient. Viruses were purified from cesium salts by dialysis. Determination of viral titers was performed by its capacity to form plaques on agarose layer.

Passage 3 MSCs were transfected with Ad5-VEGF165 and Ad5-FGF2 at 100 MOI each. The transcriptional activity of genes vegf 165 and fgf 2 were evaluated 24 h post-infection by qPCR. Specific primers, probes and nucleotide sequences were described previously (15). Serial dilutions of cDNA synthesized from mRNA of the transfected cells were used to construct a standard curve and determine the level of gene expression. The gene expression level of non-transfected cells was taken as 100%.

Matrigel™ tube-formation assay with Ad5-VEGF165 and Ad5-FGF2 co-transfected cells and non-transfected (control) cells was performed as described previously (16). Briefly, 10,000 MSCs co-transfected with Ad5-VEGF165 and Ad5-FGF2 or 10,000 non-transfected cells per well in triplicates were seeded in a 96-well plate pre-coated with 50 μl of Matrigel^® Growth Factor Reduced Basement Membrane Matrix (Cat. #356231, Becton Dickinson, USA) in DMEM/F12 media supplemented with 1% FBS. The plates were incubated at 37°C in a humidified atmosphere containing 5% CO₂ for 16 h. Tube formation was evaluated using microscopy and ImageJ software.

3D bioactive matrix was formed immediately prior to application to the wound. Fibrin glue Tissucol was used as a matrix basement. Two types of matrix were used. The first type of matrix consisted of 3,000,000 genetically modified MSCs and 1 ml of Tissucol. MSCs were co-transfected with Ad5-VEGF165 and Ad5-FGF2 at MOI 100 each the day before application. Cell suspension, thoroughly washed from media and viruses by centrifugation, was mixed with fibrin glue and uniformly applied to the wound surface. The second type of matrix was prepared from 1 ml of Tissucol and a mixture of Ad5-VEGF165 and Ad5-FGF2 viruses in amounts equivalent to those used for MSC transfection.

Statistical analysis was performed using Student's t-test in Microsoft Excel 2007 software, P ≤ 0.05 was considered statistically significant.

Adenoviruses Ad5-VEGF165 and Ad5-FGF2 were prepared using standard methods of molecular genetics and used for gene therapy as part 3D-matrix or to modify the MSCs with MOI 100. MSCs isolated from dog adipose tissue as was described previously (17) had a fibroblast-like morphology, and expressed markers of the MSCs CD10, CD71, CD44, CD105, stro-1, Thy-1, and did not express the hematopoietic markers CD34 or CD45 (Figure 1).

Biosynthesis of VEGF and FGF2 mRNA in transduced cells was confirmed after 24 h after modification of MSC with recombinant adenoviruses (Figure 2A). MSCs-Ad-VEGF165-FGF2 had a higher (70 ± 5 units/well) capacity to form capillary-like structures on Matrigel^TM as compared with intact cells (47 ± 5 units/well, Figures 2A–D), p < 0.05.

In vivo biologically active 3D matrix was tested in two morphologically similar non-healing full-thickness bite wounds measuring about 70 cm² each in the back and groin areas of a medium sized adult dog. The numerous bite wounds were received during fights between the patient and another dog. Immediately after injury, the wounds were sanitized and sutured in a veterinary clinic. The post-operative period proved complicated with extensive necrosis of the skin at the wound edges. Necrotic tissue was dissected away, as a result the closure of the wounds by matching their edges became impossible. The effect of standard practice conservative therapy was insignificant: after 2 weeks of treatment marginal necrosis was observed and granulations were almost absent. There was no wound contraction either. As the patient had not responded to standard veterinary care, treatment with the new therapy was administered. Necrotic skin was dissected and tissues were flayed until pinpoint bleeding (Figure 3).

In order to stimulate the regeneration of tissue growth around the wounds, the biologically active matrix was applied. The wound on the back was covered with matrix based on Tissucol fibrin glue containing 3 million MSCs transfected with combination of Ad5-VEGF165 and Ad5-FGF2 (MOI 100 for each construct). Allogenic MSCs had to be used due to the severity of the patients' body condition. The wound in the groin was covered with another type of matrix containing Tissucol fibrin glue, Ad5-VEGF165 and Ad5-FGF2 in amounts equivalent to those used to transduce the MSCs. Wounds were closed with occlusive dressings. To prevent wound contamination, Baytril 5% (enrofloxacin, Bayer Corporation, USA.) antibiotic was subcutaneously injected in dose 0.1 ml/kg once a day for 7 days.

On the 3rd day after 3D matrix application the dressings were removed. Juicy red granulations, edge epithelization up to 3–4 mm, wound contraction and minor exudation were observed. On the 7th day, edge epithelization of from 6–7 to 10 mm had been achieved. There were no signs of necrosis or inflammation throughout the entire healing period. The wound on dogs' back was completely healed after 1 month and the wound in the groin healed within 1.5 months with normal non-hypertrophic scar formation (Figure 4). Following successful treatment the dog was adopted and rehomed as a pet.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.